I am not weighing in on this much, just that Windows can run a lot of unix scripts. It has a unix subsystem (Windows Subsystem for Linux (WSL)). That said, PowerShell should be able to handle this w/o any issues. Process lists are pretty standard in the .NET world. I use them a lot. A quick search showed Get-Process. Either a script or PowerShell should work with similar logic. ... View more

Hi, Have you found the solution for this? I have faced something similar and solved it by changing the length of the character variables. It seems like the auto transformation occurs if the character length is huge. Though I am not sure what is the threshold value for the length. For now defining the length as 100 worked for me. Hope this helps. Also let me know if you found any other solution. Regards, Bipasha ... View more

Aha, I am so sorry, you were right. It is opposite, I have DIS 4.8 but I try to import from DIS 4.9, so it is a problem ... View more

I agrree with this but how can I fix those warnings then? ... View more

Maybe this: *Get the first and last date for each group; proc sql noprint; create table have_2 as select *, min(DateCharacter) as period_start, max(DateCharacter) as period_end from have group by ID order by ID, DateCharacter; quit; data want; set have_2; retain period; by id; *Fix the string dates to be SAS dates.; _d=cats(DateCharacter,"01"); _d_start=input(cats(period_start,"01"),yymmdd10.); _d_end=input(cats(period_end,"01"),yymmdd10.); d=input(_d,yymmdd10.); *Count the number of periods for this ID.; if first.id then period = 1; else period = period + 1; *Fix the table name by concatination.; table_name = cats(put(_d_start,date9.),"_",put(_d_end,date9.)); format d yymmdd10.; *Get rid of temporary variables; drop _d: period_start period_end; run; BTW, it is a bit confusing that you have to look at two messages to try to figure out what you want. Better to rewrite it to one instead. 🙂   ... View more

@Reeza wrote: I think you're seriously over complicating this. Are you trying to ensure that you have a full 12 month for each ID and that it's the current month, not the last month? Please confirm if this is correct and if so, I'll post a revised solution. @Reeza makes a very good point. Is it correct? You haven't given this explanation, and quite honestly, the more I think about what you said, the less clear I am on what you want. The explanation you have given doesn't seem general enough to program, except for the very specific original data set you provided, so its not difficult to provide a solution for that one original data set with 11 observations (and I have done that) but obviously you have a more general problem and we are not sure exactly what that more general problem is. ... View more

in that case, renaming using dataset option is a good idea   PROC SQL; CREATE TABLE WANT(drop=ym) AS SELECT *,Input(ym,YYMMN6.) AS YEARMONTH Format=date9. FROM Have(rename=(yearmonth=ym)); QUIT; ... View more

PROC MEANS DATA=Ranked_Have NWAY N NMISS MEAN MIN MAX STACKODS; CLASS BUCKET; VAR SCORE TARGET ; ods output summary=ranked_stats; RUN; @turcay wrote: Thank you,   I tried something but it does not work. Could you tell where should I put N, NMISS and SUM function in the following procedure?   PROC MEANS DATA=Ranked_Have NOPRINT MIN MAX N NMISS SUM ; CLASS BUCKET; VAR SCORE; VAR Target; OUTPUT OUT=RANGES SUM=Target NMISS=Target MIN=MIN MAX=MAX ; RUN;   ... View more

SAS/ACCESS to Oracle documentation lists available libname options. ... View more

Place the LOCK PORT.STATUS CLEAR; in the 1st code that you run.     Data BATCH; Length FINAL_DATE 8 ERROR 8; Infile Datalines Missover; Input FINAL_DATE ERROR ; FORMAT FINAL_DATE DATE9.; Datalines; 21244 0 21244 0 21244 0 21244 0 21244 0 21244 0 21244 0 21244 0 21244 0 21244 0 21244 0 ; LIBNAME PORT "C:\Users\...\Desktop"; Data PORT.STATUS; Start_Date="Y"; Complete_Date="01MAR2018"D; Format Complete_Date DATE9.; Run; Options dbidirectexec bufno=100 bufsize=128k; LIBNAME SQL ODBC DATAsrc=proddb SCHEMA=sas USER=user PASSWORD="*******" INSERT_SQL=YES INSERTBUFF=32767 readbuff=10000 bulkload=yes dbcommit=10000; %MACRO main_loop; PROC SQL; SELECT MAX(RESULT) INTO:errcheck FROM(SELECT *, CASE WHEN FINAL_DATE="&SYSDATE9."D AND ERROR=0 THEN 0 ELSE 1 END AS RESULT FROM BATCH); QUIT; %IF &errcheck %THEN %DO; DATA _NULL_; ABORT CANCEL; RUN; %END; %ELSE %DO; DATA _NULL_; SET PORT.STATUS; IF Start_Date="Y" OR Complete_Date="&sysdate9."d THEN ABORT CANCEL; ELSE Call system("...Batch.bat"); RUN; %END; %MEND; %main_loop; LOCK PORT.STATUS CLEAR;   Now when the above program running your .bat file your triggering the next program (Try triggering after the .bat run completed). So before the next program runs you need to check whether the lock is released or not. If the above program still running then the dataset is locked.     %PUT &SYSLCKRC.; /* Unlock if the dataset need to be updated is locked by me previously */ %macro trylock(member=, timeout=); %local starttime; %let starttime = %sysfunc(datetime());/* Current time */ /* Loop until lock is released */ /* Initially &syslckrc value is 0 */ %DO %UNTIL (&syslckrc = 0 or %sysevalf(%sysfunc(datetime()) > (&starttime + &timeout))); /* Try to lock the Dataset, If it is previously locked then &syslckrc value will be >0, Lets assume Its locked*/ lock &member NOMSG; /* *Since &syslckrc >0 it will sleep for 10 min */ %if &syslckrc ne 0 %then %let rc=%sysfunc(sleep(1000*60*10/*Minutes*/)); /* Since the &syslckrc value is >0 it will loop after sleep */ %END; %mend trylock; %trylock(member=PORT.STATUS, timeout=3600); /* If you have another dataset and need to run the rest of the query only if the lock is released. */ %trylock(member=PORT.STATUS2, timeout=3600); LIBNAME PORT "C:\Users\...\Desktop";/**/ DATA PORT.STATUS;/*It gives the LOCK error here*/ SET PORT.STATUS; Start_Date="Y"; Run; /*MAIN CODE IS HERE BUT I DO NOT TYPE IT BECAUSE IT CAN TAKE UP A LOT OF SPACE*/ DATA PORT.STATUS; SET PORT.STATUS; Complete_Date="&sysdate9."d; format Complete_Date date9.; Start_Date="N"; Run; /* Release lock on all datasets */ LOCK PORT.STATUS CLEAR; LOCK PORT.STATUS2 CLEAR; SYSLCKRC Automatic macro variable value will be 0 if the last lock statement is successful. >0 if the lock statement is not successful and <0 if warnings or note was written to log.   3550 Sec~1hr because I want to try for checking the lock for 1 hr ie. every 10 min until 1 hr. This is not a fixed number it's up to you how long you want to check for lock on the dataset depending on the previous job run time.  Note: If running in windows then remove 1000 in sleep sunction i.e  %if &syslckrc ne 0 %then %let rc=%sysfunc(sleep(60*10/*Minutes*/));     You can run the %trylock on any number of datasets you wish. It's a macro.   %trylock(member=PORT.STATUS, timeout=3600); %trylock(member=PORT.STATUS2, timeout=3600); %trylock(member=PORT.STATUS3, timeout=3600); ........................so on               ... View more

Try it. ... View more

@KachiM smallest will work, just need to look at it a bit differently:   data want; set have; array k[*} var: ; do i = 1 to dim(k); if smallest(i,of k(*)) ge 0 then do; val = smallest(i,of k(*)); leave;; end; end; run;   I will not make any claim to efficiency other than short code. Notice that i will, if a positive value is found have, the index of the smallest positive value in the array which may be helpful. ... View more

Hi @turcay,   In MIP 2.4, we introduced the concept of Risk Data Objects.  A risk data object is an object that you map to a SAS data set or data view when you configure an analysis run.  In your model groups, you can define the following types of risk data objects: cash flow, parameter matrix, and value data.   To add a cash flow data set to an analysis run, complete the following steps : On the Risk Data Objects tab ( ) in your model group, define a new cash flow object. In your user-defined logic, use the READ_SCENARIO_CF CALL routine to call the cash flow object. Add this model group to the model group map that you plan to include in your analysis run. Create a scenario run or an economic simulation run.  When prompted in the run wizard, select the cash flow data set that you want to map to the cash flow object.  The cash flow data set must be stored in the work-group-directory/input/risk_data/cash_flows directory. Please let me know if you have additional questions.   Regards,   Toshiba ... View more